Author + information
- Received October 21, 2013
- Revision received February 18, 2014
- Accepted February 25, 2014
- Published online June 3, 2014.
- Jun Pu, MD∗,†,‡,
- Gary S. Mintz, MD†,
- Sinan Biro, MSc†,
- Jin-Bae Lee, MD∗,†,
- Stephen T. Sum, PhD§,
- Sean P. Madden, PhD§,
- Allen P. Burke, MD||,¶,
- Pei Zhang, PhD||,
- Ben He, MD‡,
- James A. Goldstein, MD#,
- Gregg W. Stone, MD∗,†,
- James E. Muller, MD§,
- Renu Virmani, MD¶ and
- Akiko Maehara, MD∗,†∗ ()
- ∗Department of Medicine, Division of Cardiology, Columbia University Medical Center, New York, New York
- †Cardiovascular Research Foundation, New York, New York
- ‡Department of Cardiology, Renji Hospital, Shanghai Jiao Tong University School of Medicine, Shanghai, China
- §InfraReDx, Burlington, Massachusetts
- ||University of Maryland Medical Center, Baltimore, Maryland
- ¶CVPath Institute, Gaithersburg, Maryland
- #Division of Cardiology, William Beaumont Hospital, Royal Oak, Michigan
- ↵∗Reprint requests and correspondence:
Dr. Akiko Maehara, Columbia University Medical Center, New York-Presbyterian Hospital, Cardiovascular Research Foundation, 111 East 59th Street, 12th Floor, New York, New York 10022.
Objectives Three intravascular ultrasound (IVUS) signatures have been associated with coronary artery disease instability: echo attenuation, an intraplaque echolucent zone, and spotty calcification. The aim of this study was to investigate the substrates responsible for these IVUS signatures in a relatively large series of post-mortem human coronary samples.
Background The exact mechanisms and pathological correlates underlying echo attenuation, an intraplaque echolucent zone, and spotty calcification remain poorly understood.
Methods IVUS was compared with near-infrared spectroscopic detection of lipid core plaque and histopathology in 2,294 vessel segments from 151 coronary specimens from 62 patients at necropsy using the modified American Heart Association classification.
Results IVUS detected echo-attenuated plaques in 18.3% of segments, echolucent plaques in 10.5% of segments, and spotty calcification in 14.4% of segments. Histopathologically, 91.4% of echo-attenuated plaques corresponded to either a fibroatheroma (FA) with a necrotic core (NC) or pathological intimal thickening with a lipid pool; almost all segments with superficial echo attenuation indicated the presence of an FA with an advanced NC. Echolucent plaques indicated the presence of a relatively smaller lipid or NC compared with echo-attenuated plaques (thickness: 0.51 mm [interquartile range (IQR): 0.35 to 0.64 mm] vs. 0.70 mm [IQR: 0.54 to 0.92 mm] [p < 0.001]; arc: 74.5° [IQR: 59.0° to 101.0°] vs. 90° [IQR: 70.0° to 112.0°] [p < 0.001]), although 82.8% of superficial echolucent zones indicated an NC-containing FA. IVUS spotty calcification, especially when superficial in location (72.6%), was often associated with an FA with calcium deposits and had smaller arcs of calcium in the setting of FA compared with fibrocalcific plaques (37.5° [IQR: 23.0° to 53.0°] vs. 59.0° [IQR: 46.0° to 69.0°]; p < 0.001). Comparisons between IVUS and near-infrared spectroscopy revealed that echo-attenuated plaques contained the highest probability of near-infrared spectroscopy–derived lipid core plaque, followed by echolucent plaques and spotty calcifications.
Conclusions This study demonstrated that echo-attenuated plaque, especially superficial echo attenuation, was the most reliable IVUS signature for identifying a high-risk plaque (i.e., an FA containing a large NC).
- attenuated plaque
- intravascular ultrasound
- near-infrared spectroscopy
- spotty calcification
Intravascular ultrasound (IVUS) has played an important role in understanding the pathology and treatment of atherosclerotic disease in humans. Several grayscale IVUS features have been associated with either clinical instability or a high risk for cardiovascular events in patients with coronary artery disease undergoing percutaneous coronary intervention. These include atherosclerotic plaque with ultrasonic attenuation (echo-attenuated plaque) (1,2), an intraplaque echolucent zone (echolucent plaque) (3), and scattered spotty calcification (4,5). To date, however, the exact mechanisms or pathological correlates underlying these IVUS signatures and the associated risk for ischemic events remain poorly understood. More recently, near-infrared spectroscopy (NIRS) has been developed to detect lipid core plaque (LCP) (6,7); NIRS has been fused with IVUS in the first combined imaging system (8). In the present ex vivo study, we investigated the substrates responsible for echo-attenuated plaques, echolucent plaques, and IVUS spotty calcification on the basis of the modified American Heart Association histological classification scheme in a relatively large series of post-mortem human coronary samples.
Human coronary specimens were obtained over a 2-year period from autopsied patients. With approval from the institutional review board, hearts were acquired from the National Disease Research Interchange (Philadelphia, Pennsylvania) or the International Institute for the Advancement of Medicine (Edison, New Jersey), which also provided information regarding age, sex, medical history, and cause of death. The following terms were designated as indicating a cardiovascular cause of death: “myocardial infarction,” “cerebrovascular accident,” “cardiac arrest,” and “other cardiovascular.” Hearts were received within 48 h of death, maintained on ice at 4°C, and imaged within 96 h of death. The major coronary arteries were harvested after in situ angioscopic screening to exclude occluded segments impassable by the IVUS catheter. Side branches were ligated to prevent loss of blood during perfusion; adventitial fat surrounding the arterial segments was kept intact. Each coronary specimen was mounted in a unique custom fixture with vertical guideposts at 2-mm intervals to be used as reference points when comparing imaging with histological studies. Both ends of the arterial segment were attached to Luer connectors that allowed fluid flow and catheter entry. A varistaltic pump (Manostat, Barnant Corporation, Barrington, Illinois) supplied pulsatile flow at 60 cycles/min and a flow rate of approximately 130 ml/min, with pressure inside the coronary artery maintained at physiological levels (80 to 120 mm Hg) at a body temperature of 37.0°C. IVUS, NIRS, and histopathologic analyses were performed without knowledge of the findings obtained from the other 2 methods.
IVUS image acquisition and analysis
An Atlantis SR Pro 40-MHz catheter attached to an iLab system (Boston Scientific Corporation, Fremont, California) was advanced along a 0.014-inch guidewire through the coronary specimen mounted in the fixture. IVUS imaging was performed using motorized pullback at 0.5 mm/s to include proximal and distal Luer connectors. Image data were archived onto a CD-ROM and sent to an independent IVUS core laboratory (Cardiovascular Research Foundation, New York, New York) for off-line analyses. Every IVUS frame was matched to its comparable histopathologic slice using vessel shape, side branches, perivascular structures, and distances from the Luer connectors.
IVUS analyses were performed using validated planimetry software (echoPlaque, INDEC Medical Systems, Santa Clara, California). Quantitative analyses were performed according to criteria from the American College of Cardiology consensus statement on IVUS (9). Echo-attenuated plaque, echolucent plaque, and IVUS spotty calcification were defined as previously published (1–5,10).
Echo-attenuated plaque was identified by the absence of the ultrasound signal behind plaque that was either hypoechoic or isoechoic to the reference adventitia but contained no bright calcium (10). The arc of attenuation was measured in degrees with a protractor centered on the lumen. The interobserver variability of this arc of attenuation measurement was 4.9 ± 2.8°. The location of attenuation was defined as superficial (leading edge of attenuation closer to the lumen than to the adventitia) or deep (leading edge of attenuation closer to the adventitia than to the lumen).
Echolucent plaque contained an intraplaque zone of absent or low echogenicity (lower than that of the reference adventitia) surrounded by tissue of greater echodensity (3,10). The arc of echolucent zone was measured in degrees with a protractor centered on the lumen. The interobserver variability of the arc of echolucent zone was 5.8 ± 3.4°. The location of the echolucent zone was defined as “superficial” if the leading edge of the echolucent zone was closer to the lumen than to the adventitia or “deep” if the leading edge of the echolucent zone was closer to the adventitia than to the lumen (3).
Calcification was brighter than the reference adventitia and caused acoustic shadowing with or without reverberations (9). The arc of calcium was measured with a protractor centered on the lumen. The interobserver variability of the arc of calcium measurement was 4.1 ± 1.9°. Spotty calcification contained small calcium deposits within arcs of <90° (4,5). The location of calcium was superficial (closer to the lumen than to the adventitia), deep (closer to the adventitia than to the lumen), or mixed (both superficial and deep) (9).
All IVUS determinations were based on the observations of 2 independent experienced reviewers (J.P. and A.M.) who were blinded to other data; lesions for which these 2 observers disagreed were excluded. Interobserver and intraobserver variability yielded good concordance for the diagnosis of echo-attenuated plaque (κ = 0.90 and κ = 0.93, respectively) and spotty calcification (κ = 0.93 and κ = 0.96, respectively), with moderate concordance for the diagnosis of echolucent plaque (κ = 0.74 and κ = 0.77, respectively).
Near-infrared spectroscopic image acquisition and analysis
Using the same protocol as for IVUS imaging, a 3.2-F InfraReDx (Burlington, Massachusetts) near-infrared spectroscopic catheter was advanced into the distal coronary vessel (6,10). Automated mechanical pullback was performed at a speed of 0.5 mm/s and 240 rpm. Raw spectra were acquired at a rate of 40 Hz (1 spectrum every 25 ms). Spatial filtering and image processing of the raw data produced an image with data points every 0.1 mm and every 1°.
During catheter pullback, the probability of LCP was displayed as a “chemogram,” a digital color-coded map of the location and intensity of lipid viewed from the luminal surface, with the x-axis indicating the pullback position in millimeters (every 0.1 mm) and the y-axis indicating the circumferential position in degrees (every 1°) as if the coronary vessel had been split open along its longitudinal axis. Spectroscopic information at each pixel was transformed into a probability of LCP, which was then mapped to a 128-point (7-bit) red-to-yellow color scale with the lowest probability of lipid shown as red and the highest probability of lipid shown as yellow. The “block chemogram” was a summary metric that displayed the probability that was present for all measurements using the top 10th percentile pixel information (i.e., the 90th-percentile value) of the corresponding 2-mm chemogram segment. If the probability of the top 10th percentile was ≥0.98, the entire block was yellow.
Near-infrared spectroscopic image analyses were performed off-line using in-house, MATLAB-based software programmed at the Cardiovascular Research Foundation (10). The presence of LCP required 1 yellow block. Yellow pixels (pixels above the preset threshold for the detection of LCP) within the analyzed segment were divided by all pixels in the chemogram to generate the lipid core burden index (LCBI) per mille.
Histopathologic analysis and definitions
After the IVUS and near-infrared spectroscopic images were obtained, segments were pressure fixed in formalin. To obtain histopathologic sections corresponding to the IVUS and near-infrared spectroscopic cross sections, vessels were divided into consecutive 2-mm–long arterial segments using the fixture’s guideposts. Histological preparation and analysis were performed at the CVPath Institute (Gaithersburg, Maryland). Histopathologic slides 5-μm thick were prepared from the distal side of each 2-mm block and stained with hematoxylin and eosin and Russell-Movat’s pentachrome, as previously described (11). With Movat pentachrome, elastic fibers stain black, collagen fibers stain yellow, proteoglycan stains green, and smooth muscle cells stain red. Each histopathologic slide was digitized using a standard microscope. Histological features were classified using the Virmani-modified American Heart Association plaque classification scheme on the basis of the agreement of 2 pathologists from the CVPath Institute (A.P.B. and R.V.), who were blinded to the IVUS and near-infrared spectroscopic findings (12–15).
In brief, pathological intimal thickening (PIT) was characterized by the formation of extracellular non-necrotic lipid pools within a structured matrix typically adjacent to the media. Early fibroatheromas (FAs) had early necrotic cores (NCs) characterized by small cholesterol clefts and cellular debris with structured matrix. Late FAs had late NCs characterized by increased cellular debris and cholesterol clefts but complete degradation of extracellular structured matrix. Thin-cap FAs had large late NCs covered by thin fibrous caps (<65 μm). Examples are shown in Online Figure 1. Nonlipidic plaques were lesions lacking NCs or lipid pools, including adaptive intimal thickening, smooth muscle cell–rich plaque, bland fibrous plaque, intimal xanthoma, calcified fibrous plaque, and calcified nodules. Using custom software (IPLab, Scanalytics, Rockville, Maryland), contours were drawn on each image to identify fiduciary markers retained by the tissue preparation inking scheme, luminal boundary, lipid pools at sites of PIT, and NC in FA. Quantitative assessment of the thickness, arc, and size of lipid/NC was determined using an automatic computer-assisted technique, and the size of lipid/NC was expressed as the percent cross-sectional area of the total plaque, as previously described (11). On the basis of the report that the mean size of NCs was 23% in thin-cap FAs (15), we defined “large” lipid/NC as >20%.
Data analyses were performed using SPSS for Windows version 15.0 (IBM, Armonk, New York) and SAS version 9.13 (SAS Institute, Cary, North Carolina) (16). Categorical data are expressed as absolute values and percents and were compared using chi-square or Fisher exact tests, and continuous data are reported as median (interquartile range [IQR]) and were compared using Kruskal-Wallis/Wilcoxon rank sum tests. For segment-level analyses, a model with a generalized estimating equation approach was used to compensate for any potential cluster effect of multiple segments in the same subject. Intraobserver and interobserver variability was assessed using Cohen’s kappa statistics, which correct for the chance of simple agreement and account for systematic observer bias. A kappa value of 0.81 to 1.0 indicates good agreement, and a value of 0.61 to 0.80 indicates moderate agreement. The correlation of the IVUS measurements with histologic findings was calculated using the Spearman correlation coefficient. p values were adjusted using the generalized estimating equation method for repeated measures. Inferential statistical tests were conducted at a significance level of 0.05. The authors had full access to and take full responsibility for the integrity of the data. All authors have read and agree to the report as written.
Demographic data and lesion population
The median lengths of the coronary specimens imaged were 46 mm (IQR: 42 to 50 mm) in the left anterior descending coronary artery, 46 mm (IQR: 38 to 52 mm) in the left circumflex coronary artery, and 50 mm (IQR: 42 to 58 mm) in the right coronary artery. From these imaging runs and subsequent histopathologic analyses, a total of 2,294 2-mm–long segments from 151 coronary specimens from 62 autopsied hearts were used for the present study. Of these, 1,234 segments (53.8%) were classified as plaques containing NCs or lipid pools: 788 (34.4%) were FAs containing NCs, and 446 (19.4%) were PIT containing lipid pools. Donors’ median age was 64 years (IQR: 51 to 77.5 years; range 39 to 83 years), 66.1% were men, and 77.4% had cardiovascular causes of death (Table 1).
Overall, 419 segments (18.3%) contained echo attenuation; the median arc of attenuation was 80° (IQR: 60° to 104°; range 29° to 360°), and 99.3% of the cross sections demonstrating attenuated plaque had >30° of attenuation (Fig. 1). On pathological analyses, 91.4% of segments with echo attenuation contained NCs or lipid pools at the site of echo attenuation (Fig. 2): 67.1% were FAs containing NCs (21.2% early NCs and 45.8% late NCs), and 24.3% were non-FAs (i.e., PIT) containing lipid pools.
The prevalence of echo attenuation was highest in FAs with late NCs (39.5%), followed by FAs with early NCs (28.1%) and PIT with lipid pools (22.8%) (p < 0.001). The prevalence of echo attenuation also increased with the size of the NC or lipid pool (Figs. 3A to 3C). When histopathology was considered the gold standard, IVUS echo-attenuated plaque had a sensitivity of 56.2% and a specificity of 94.7% for the detection of a large lipid/NC (Table 2). Histopathologically, the arc of lipid/NC that produced echo attenuation measured 90° (IQR: 70° to 112°; range 31° to 360°), and the arc of IVUS attenuation was significantly correlated with the arc of histopathologic lipid/NC (Spearman ρ = 0.637, p < 0.001) (Fig. 1D). The median thickness of lipid/NC that produced ultrasonic attenuation was 0.70 mm (IQR: 0.54 to 0.92 mm; range 0.30 to 2.57 mm). Echo attenuation in the absence of an NC or a lipid pool was seen in only 8.6%, mainly at a histological branch point (6.7%) or in association with a calcium deposit (1.9%).
Overall, 52.3% of echo attenuation was superficial in location and 47.7% was deep in location. Attenuation was almost always (96.9%) concordant to the maximal plaque plus media thickness. When classified by location, 96.3% of superficial attenuation indicated an FA containing an NC; none indicated PIT with a lipid pool (Fig. 3D). Conversely, 52.8% of deep attenuation indicated PIT containing a lipid pool, whereas only 31.5% indicated an FA containing an NC (Fig. 3E). Compared with plaques with deep attenuation, those with superficial attenuation were more often associated with an FA or an FA with a late NC (p < 0.001 for both) (Fig. 3F). PIT containing a lipid pool only caused attenuation closer to the adventitia than to the lumen (i.e., deep) regardless of the size of lipid pool and never attenuation closer to the lumen than to the adventitia.
Overall, 241 segments (10.5%) contained IVUS echolucent zones. The median arc of the echolucent zone was 70° (IQR: 55° to 88°; range 31° to 162°). The frequency distribution of the arc of the echolucent zone is shown in Figure 4A.
On pathological analyses, 76.8% of segments with echolucent zones contained NCs or lipid pools; 48.1% were FAs with NCs (28.2% late NCs and 19.9% early NCs), and 28.6% were PIT with lipid pools. Examples are shown in Figure 5. The prevalence of echolucent plaques was highest when the lipid/NC represented 20% to 40% of plaque area, with no significant difference in the prevalence of echolucent plaques between NC plaques (14.7%) and lipid pool plaques (15.5%) (p = 0.723), and between FAs with early NCs (15.1%) and those with late NCs (14.4%) (p = 0.784) (Figs. 4C to 4E). When histopathology was considered the gold standard, IVUS echolucent plaque had a sensitivity of 20.5% and a specificity of 90.4% for the detection of a large lipid/NC (Table 2). The median histopathologic arc of lipid/NC that produced an echolucent zone was 74.5° (IQR: 59.0° to 101.0°; range 34.0° to 188.0°); the arc of the echolucent zone correlated with the arc of histopathologic lipid/NC (Spearman ρ = 0.556, p < 0.001) (Fig. 4B). The median thickness of lipid/NC that produced an echolucent zone was 0.51 mm (IQR: 0.35 to 0.64 mm; range 0.29 to 1.79 mm). In 23.2% of segments, echolucent zones did not indicate histopathologic NCs or lipid pools but were associated with intraplaque hemorrhage (9.5%), nonlipidic plaque containing loose collagen (8.3%), or artifacts of the pathological preparation process (5.4%).
Overall, 48.1% of echolucent zones were superficial in location and 51.9% were deep in location. The echolucent zone was almost always (90.9%) concordant to the maximal plaque plus media thickness. When classified by location, most (82.8%) superficial echolucent zones indicated FAs containing NCs (56.0% late NCs and 26.7% early NCs); none indicated PIT with a lipid pool. Deep echolucent zones mostly indicated PIT containing lipid pools: 55.2% were PIT, and only 16.0% were FAs with NCs. Compared with deep echolucent zones, superficial echolucent zones less often represented PIT (0% vs. 55.2%, p < 0.001) and more often represented FAs (82.8% vs. 16.0%, p < 0.001) and FAs with late NCs (56.0% vs. 2.4%, p < 0.001).
Compared with echo-attenuated plaques, echolucent plaques had a significantly smaller thickness (p < 0.001) (Fig. 6A) and arc (p < 0.001) (Fig. 6B) of histological lipid/NC. In addition, echolucent plaques less often represented FAs (p < 0.001) (Fig. 6C) or FAs with late NCs (p < 0.001) (Fig. 6D) compared with echo-attenuated plaques.
The cut point for thickness of lipid/NC that differentiated attenuated plaques from echolucent plaques measured 0.60 mm, with a sensitivity of 67% and a specificity of 71% (area under the curve = 0.74). The cut point for arc of lipid/NC that differentiated attenuated plaques from echolucent plaques measured 75.5°, with a sensitivity of 66% and a specificity of 59% (area under the curve = 0.64).
IVUS detected calcification in 735 segments (32.0%). The median arc of calcium was 93.0° (IQR: 54.0° to 140.0°; range 9.0° to 360.0°), and 44.9% of these segments had <90° arc of calcium with a spotty pattern. Overall, 69.9% of all IVUS-detected calcifications were superficial in location, 10.2% were deep in location, and 19.9% were mixed. On pathological analyses, 54.6% of segments with IVUS calcification were calcified fibrous plaques, 40.4% were FAs with calcium deposits, and 5.0% were calcified nodules (Fig. 7). IVUS did not detect calcium in 14.8% of segments containing histopathologic calcium; reasons included deep calcium hidden behind a large NC that produced echo attenuation (5.4%) and microcalcium deposits (9.4%).
In the setting of IVUS-detected spotty calcification, 62.4% were FAs with calcium deposits and 32.7% were fibrocalcific plaques; the median arc of calcium was 52.0° (IQR: 37.0° to 70.0°; range 9° to 89°), with smaller arcs in the setting of FA compared with fibrocalcific plaques (p < 0.001) (Fig. 8A). The frequency distribution of the arc of spotty calcification in FAs and fibrocalcific plaques is shown in Figure 8B. When histological results were considered the standard, IVUS spotty calcification had a specificity of 71.7% and a sensitivity of 69.4% for the detection of an FA with calcium deposits (Table 2). In addition, plaques with IVUS spotty calcifications were more often FAs, compared with IVUS extensive calcification (p < 0.0001) (Fig. 8C).
Overall, 67.6% of IVUS spotty calcifications were superficial in location, 13.0% were deep in location, and 19.4% were mixed; 76.3% of IVUS spotty calcifications were concordant to the maximal plaque plus media thickness. When classified by location, 72.6% of IVUS superficial spotty calcifications were seen in FAs with calcium deposits (45.3% late NCs and 27.3% early NCs). Conversely, 67.4% of IVUS deep spotty calcifications were seen in fibrocalcific plaques without NCs, and only 32.6% of IVUS deep spotty calcifications were seen in FAs. In addition, deep IVUS spotty calcifications, when seen in FAs, were uncommonly associated with FAs containing large NCs (12.5% containing large NCs [size >20%] and 87.5% containing smaller NCs) because large NCs often produced echo attenuation, hiding deeper calcific deposits (Fig. 7E).
IVUS versus NIRS findings
On NIRS analyses, 83.5% of echo-attenuated segments contained LCP, as identified by the presence of the yellow block chemogram (Fig. 9A). Compared with echo-attenuated plaques, echolucent plaques were associated with less NIRS-derived LCP (p < 0.001) (Fig. 9B) and a smaller NIRS-derived LCBI (p < 0.001) (Fig. 9C). In plaques with calcifications, IVUS spotty calcifications were associated with more NIRS-derived LCP (47.9% vs. 15.6%, p < 0.001) and a larger NIRS-derived LCBI (95.3 [IQR: 30.6 to 183.6] vs. 44.5 [IQR: 1.4 to 129.4]; p < 0.001) than IVUS extensive calcification (Fig. 10). In the setting of IVUS spotty calcification, the arc of calcification was negatively correlated with the NIRS-derived LCBI (Spearman ρ = −0.417, p < 0.001) (Online Fig. 2).
Association between plaque characteristics and cardiovascular events
As shown in Table 3, patients in this autopsy series who died of cardiovascular causes had more echo-attenuated plaques (p = 0.025) and more superficial echo-attenuated plaques (p = 0.023) compared with those who died of noncardiovascular causes. The frequency of echo-attenuated plaque involving multiple (2 or 3) epicardial vessels was 47.9% in cardiovascular death versus 21.4% in noncardiovascular death (p = 0.077). Although there were no significant differences in the frequency of echolucent plaques (p = 0.346) and spotty calcifications (p = 0.120) between the 2 groups, superficial echolucent plaques (p = 0.052) and superficial spotty calcifications (p = 0.049) tended to be more frequent in patients who died of cardiovascular causes.
The novel findings of the present study include the following: 1) IVUS attenuation indicated the presence of a large NC or lipid pool, and the closer the attenuation was to the lumen, the more likely it indicated the presence of an advanced NC; 2) an intraplaque echolucent zone indicated the presence of a relatively smaller NC or lipid pool, especially compared with echo attenuation, although most superficial echolucent zones indicated NC-containing FAs; and 3) IVUS spotty calcification, especially superficial in location, was often associated with an FA with calcium deposits and had smaller arcs of calcium in the setting of FAs compared with fibrocalcific plaques. These findings were supported by a comparison of patients who died of cardiovascular causes and those who died of noncardiovascular causes.
Ultrasonic attenuation in the absence of associated calcium (i.e., echo-attenuated plaque) was a novel IVUS imaging finding that has not been described in the existing IVUS guidelines from the American College of Cardiology or the European Society of Cardiology (9,17). Recent clinical studies have shown that echo-attenuated plaques were associated with ST-segment elevation myocardial infarction and periprocedural myonecrosis or no-reflow in patients with coronary artery disease undergoing percutaneous intervention (1,2,18). However, histopathologic analyses of a small number of specimens yielded inconsistent results, and echo attenuation has been variously related to microcalcification, hyalinized fibrous tissue, cholesterol crystals, or organized thrombus (19–22). Hara et al. (19) reported a patient who died of acute myocardial infarction and suggested that echo attenuation was caused by the presence of stratified microcalcification. Tsunoda et al. (20) examined material retrieved during atherectomy of an attenuated plaque and reported hyalinized fibrous plaque with expansive remodeling. Kimura et al. (21) examined retrieved atherectomy specimens from 30 attenuated plaques and found advanced atherosclerosis consisting predominantly of cholesterol clefts, macrophage infiltration, and microcalcification. Based on the modified American Heart Association histological classification in a relatively large series of postmortem human coronary samples, the present study demonstrated that IVUS attenuation was indicative of either an FA containing a large NC or PIT with a large lipid pool, and the prevalence of echo attenuation increased with the size of the NC or lipid pool.
The size of the lipid/NC has been significantly associated with the likelihood of plaque rupture (12,23). Davies et al. (24) estimated that when >40% of the plaque consists of lipid/NC, an atheroma is at high risk for rupture. In the present study, the prevalence of ultrasonic attenuation was highest when the lipid/NC represented >40% of plaque area. Furthermore, superficial echo attenuation was seen almost exclusively in the setting of an NC-containing FA; and the closer to the lumen, the more likely the attenuation indicated an advanced NC. Conversely, attenuation from a lipid pool in PIT always appeared closer to the adventitia than to the lumen, especially near the intimal-medial border. It is believed that PIT, characterized by the formation of extracellular non-necrotic lipid pools within the plaque, is the first progressive lesion of atherosclerosis and represents a precursor lesion to FA (14). An NC within an FA is thought to arise from macrophage infiltration of lipid pools and cell death (“early” NC), followed by secondary necrosis (“late” NC) whereby cholesterol clefts, microcalcification, and intraplaque hemorrhage likely contribute to its enlargement (13). Previous studies have speculated that microcalcification and cholesterol crystals within an advanced plaque are responsible for ultrasonic attenuation by reflecting and dispersing ultrasonic wave signals (1,21). However, in the present study, the appearance of echo attenuation in PIT without microcalcification or cholesterol crystals suggests that lipid pool alone can cause ultrasonic attenuation and that the presence of other plaque components (i.e., cholesterol crystals or microcalcifications in FA) was not necessary for the appearance of this IVUS signature. However, the more advanced the atherosclerotic plaque, the more complex and heterogeneous the plaque components that contribute to enhancing attenuation of ultrasound (i.e., closer to the lumen).
One prospective clinical study with 2-year follow-up found an increased risk for acute coronary events in patients with IVUS superficial (but not deep) echolucent zones within coronary plaques (3). In the present study, an intraplaque echolucent zone indicated the presence of an NC or a lipid pool that was smaller than that seen with echo-attenuated plaques. Moreover, most superficial echolucent zones were seen in the setting of an FA containing an NC, whereas deep echolucent zones indicated PIT containing a lipid pool. The exact mechanism for grayscale IVUS echolucent zones remains debated (25). It has been postulated that echolucency should be a sign of homogeneous tissue (lipid pool), whereas NC tissue is heterogeneous with multiple tissue interfaces and therefore should be hyperechoic but not hypoechoic. However, the results of the present histopathologic study show that echolucent areas actually indicated areas of high lipid content within either a preserved (lipid pool) or degraded (NC) extracellular matrix. In addition, the present data also indicate significant limitations when echolucent zones alone are used to identify lipid-rich plaques. Compared with echolucent plaques, echo-attenuated plaques had a significantly greater thickness and arc of lipid/NC on histopathology. In addition, the prevalence of echolucent plaques was highest when the lipid/NC represented 20% to 40% of plaque area, whereas the prevalence of echo-attenuated plaques was highest in plaques with a percent core of >40%. Thus, a relatively small area of lipid/NC most likely produced an echolucent area on IVUS, whereas a larger lipid/NC produced echo attenuation.
Although extensively calcified lesions are assumed to be clinically quiescent, spotty calcification has been associated with an increased incidence of ischemic cardiovascular events (4,5,26). Recent clinical studies have shown that: 1) the pattern of calcification is different in patients with acute coronary syndromes compared with those with stable angina (4,27); 2) IVUS-identified spotty calcification is more likely to be found in culprit lesions in patients with myocardial infarctions than in patients with stable angina (4); 3) IVUS spotty calcification identifies patients with an active state that has accelerated disease progression (5); 4) ruptured coronary plaques are associated with spotty calcification, particularly in deep locations, and the number of deep calcium deposits was an independent predictor of culprit plaque ruptures in patients who had acute coronary syndromes (27); and 5) IVUS superficial spotty calcification patterns are associated with very late stent thrombosis after bare-metal stent implantation (28). However, studies on the histological plaque composition of IVUS-detected spotty calcified lesions are lacking.
The present histopathologic study supports the association between spotty calcification and unstable plaque. First, IVUS detected spotty calcium more often in an FA than in fibrocalcific plaque. Second, in the setting of IVUS spotty calcification, FA with calcium deposits was characterized by smaller arc of calcium deposits versus fibrocalcific plaques. Conversely, the present histopathologic study does not support the concept that IVUS-detected deep spotty calcium is associated with unstable plaque characteristics that predict future rupture. Rather, deep calcium behind a large NC was often not detected by IVUS because of echo attenuation produced by the overlying NC. Compared with the gold standard of histopathology, IVUS deep spotty calcifications were more likely to be seen in plaques without large NCs. It is more likely that deep spotty calcium behind a large NC was detected by IVUS at the bottom of a cavity after, rather than before, plaque/NC rupture.
Pathological studies have demonstrated that calcium is a frequent feature of ruptured plaques underlying sudden cardiac death and that a few scattered small calcium deposits are often present in the fibrous cap of an FA (29–31). The development of scattered small calcium around the NC appears to result from the induction of osteogenic changes by inflammatory factors and oxidized lipids in NC (32,33). The contribution of spotty calcification to plaque vulnerability is likely related to mechanical instability introduced by coherent focal calcium deposits at the interface with noncalcified plaque components (30,31).
First, this study was autopsy based, in comparison with the gold standard of histopathology. Although vessels were studied as soon as possible, such ex vivo study was limited by possible tissue changes induced by histopathologic preparation and lacking the dynamic motion and tortuous anatomy of the coronary vasculature.
Second, we could not assess very severe stenoses not passable by the IVUS or NIRS catheter. These segments were excluded by in situ angioscopic screening before examination.
Third, IVUS qualitative analysis may be subject to poor reproducibility (34). Qualitative assessment of echogenicity is performed on the cross-sectional view using the adventitia as a reference; however, the echogenicity of the adventitia may change from frame to frame.
Fourth, the present observations might be valid only for the 40-MHz IVUS catheter, because different IVUS frequencies have different penetrations that might affect qualitative plaque characteristics, especially attenuation. Similarly, ultrasound attenuation might be affected by the position of the IVUS catheter relative to the maximal plaque thickness.
Finally, hearts were obtained from 2 research tissue procurers and may not be representative of the general population or a specific subset of patients.
The present study in a relatively large series of human coronary autopsy specimens from victims of both cardiovascular and noncardiovascular deaths suggests that standard grayscale IVUS can provide more ultrasound information on plaque characterization than previously considered. Echo attenuation indicates a relatively larger lipid/NC than an echolucent zone, and IVUS spotty calcification is closely related to the presence of an NC, also indicating plaque instability. Those ultrasonic signatures, especially when superficial in location, are likely associated with an advanced FA with an NC and, thus, greater biologic instability than when seen to be deep in location.
The authors thank Jennifer B. Lisauskas, MS, for her technical contributions to this report.
For supplemental figures and their legends, please see the online version of this article.
Dr. Pu has received research grants from the National Natural Science Foundation of China (81170192/81270282) and Boston Scientific. Dr. Mintz has received consulting fees from Boston Scientific Corporation, InfraReDx, and Volcano Corporation; and has received research grant support from Volcano Corporation. Mr. Biro is a current employee of Volcano Corporation. Drs. Sum, Madden, and Muller are current employees of InfraReDx. Dr. Madden is a shareholder in InfraReDx. Drs. Burke and Virmani were consultants to InfraReDx for histological studies. Dr. Goldstein is a consultant to and an equity owner in InfraReDx. Dr. Stone is a consultant to Boston Scientific Corporation, Volcano Corporation, and InfraReDx. Dr. Muller has equity ownership in InfraReDx. Dr. Virmani is a consultant for and/or has received honoraria and/or research support from Abbott Vascular, Atrium, Biosensors International, Boston Scientific Corporation, CeloNova, Cordis Corporation, GlaxoSmithKline, Kona Medical, Lutonix, Medtronic, Microport Medical, Orbus Neich, ReCor Medical, SINO Medical Technology, Terumo Corporation, Merck, 480 Biomedical, and W.L. Gore. Dr. Maehara is a consultant to and has received research and grant support from Boston Scientific Corporation; and has received lecture fees from St. Jude Medical and Volcano Corporation. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- interquartile range
- intravascular ultrasound
- lipid core burden index
- lipid core plaque
- necrotic core
- near-infrared spectroscopy
- pathological intimal thickening
- Received October 21, 2013.
- Revision received February 18, 2014.
- Accepted February 25, 2014.
- American College of Cardiology Foundation
- Wu X.,
- Mintz G.S.,
- Xu K.,
- et al.
- Yamagishi M.,
- Terashima M.,
- Awano K.,
- et al.
- Ehara S.,
- Kobayashi Y.,
- Yoshiyama M.,
- et al.
- Kataoka Y.,
- Wolski K.,
- Uno K.,
- et al.
- Gardner C.M.,
- Tan H.,
- Hull E.L.,
- et al.
- Kini A.S.,
- Baber U.,
- Kovacic J.C.,
- et al.
- Bourantas C.V.,
- Garcia-Garcia H.M.,
- Naka K.K.,
- et al.
- Mintz G.S.,
- Nissen S.E.,
- Anderson W.D.,
- et al.
- Pu J.,
- Mintz G.S.,
- Brilakis E.S.,
- et al.
- Burke A.P.,
- Kolodgie F.D.,
- Farb A.,
- Weber D.,
- Virmani R.
- Virmani R.,
- Kolodgie F.D.,
- Burke A.P.,
- Farb A.,
- Schwartz S.M.
- Kolodgie F.D.,
- Burke A.P.,
- Nakazawa G.,
- Virmani R.
- Pu J.,
- Yuan A.,
- Shan P.,
- et al.
- Di Mario C.,
- Gorge G.,
- Peters R.,
- et al.
- Shiono Y.,
- Kubo T.,
- Tanaka A.,
- et al.
- Tsunoda T.,
- Hara H.,
- Kunimasa T.,
- et al.
- Kimura S.,
- Kakuta T.,
- Yonetsu T.,
- et al.
- Davies M.J.,
- Richardson P.D.,
- Woolf N.,
- Katz D.R.,
- Mann J.
- Thim T.,
- Hagensen M.K.,
- Wallace-Bradley D.,
- et al.
- Watabe H.,
- Sato A.,
- Akiyama D.,
- et al.
- Falk E.,
- Nakano M.,
- Bentzon J.F.,
- Finn A.V.,
- Virmani R.
- Johnson R.C.,
- Leopold J.A.,
- Loscalzo J.
- Aikawa E.,
- Nahrendorf M.,
- Figueiredo J.L.,
- et al.
- Flammer A.J.,
- Gössl M.,
- Widmer R.J.,
- et al.